Method for measuring net internal volume of a receptacle containing an unknown volume of residual liquid

Abstract

A method for measuring the net internal volume of test receptacles each containing an unknown but variable volume of a liquid, such as in testing automatic transmissions containing residual amounts of transmission fluid, to determine the exact volume of fluid needed to be added to fill the transmission to the correct level. The method places a pressurized reference vessel in communication with the internal spaces of a sealed transmission upon opening of a valve allowing the pressure to fall by expansion from the reference vessel into the transmission. The resulting reduced pressure is initially sensed shortly after the valve is opened, and subsequently after a more prolonged interval, allowing the pressure to be substantially stabilized. The receptacle net internal volume is then determined from the relationship between the preliminary pressure in the reference vessel and the stabilized final pressure by application of Boyle's law, and the volume of liquid to be added calculated therefrom. The difference between the initial and subsequent pressure readings sensed after the valve is opened is used to generate a correction applied to the calculated volume to offset the error created by an outflow through any small leaks which might exist.

Description

BACKGROUND OF THE INVENTION

1. Field of the Invention

This invention concerns measurement of internal volumes, and more particularly measurement of the net internal volume of successive test pieces containing unknown amounts of a liquid. Such measurements are useful in the determination of the amount of liquid to be added to bring the liquid to a prescribed level, as in filling transmissions during automobile assembly operations which transmissions contain an unknown residual volume of transmission fluid from previous testing operations.

2. Description of the Prior Art

It has heretofore been known to measure the internal volume of a receptacle by connecting a reference vessel pressurized with compressed air to the unpressurized interior of the receptacle, and allowing the compressed air to expand into this interior space of the receptacle. The final pressure level thereafter subsisting in the space comprised of the interior of both the receptacle and the reference vessel is related to the beginning pressure in the same proportion that the volume of the reference vessel bears to the total volume of the connected reference volume and receptacle, as stated by Boyle's Law. Since the volume of the reference vessel is known as well as the "dry" volume of the receptacle, the precise net volume of the receptacle with an unknown volume of liquid may be computed.

Such a method assumes that no leaks exist, since the presence of such leaks would invalidate the calculations due to the loss of pressurized fluid during the test, causing the final pressure to be lower than if no leaks existed.

As noted, the above basic method has been applied to measure the volume of residual transmission fluid in transmissions during automotive assembly operations. In applying the method, the various outlets are sealed, i.e., tachometer, vent, cooler lines, etc., and a connection made to the reference tank, typically to the filler tube.

Imperfect sealing of the openings sometimes occurs during this procedure, and slight leakage may also occur through the transmission housing joints or at various other points such as past the transmission shaft seals.

The present invention enables the effects of such leaks on the volume measurement to be corrected to produce accurate measurement despite the presence of small leaks.

SUMMARY OF THE INVENTION

According to the present invention, the presence of a small leak is detected by initially reading the pressure level shortly after the reference vessel and the receptacle are placed in communication with each other, as well as subsequently after the pressure level existing in the system has stabilized.

The difference between the initial and subsequent pressure value readings has been determined by the present inventors to be the same for each type of test piece, if no leak or a leak of constant flow rate is present. That is, variations in the stabilized final pressure occur because of the presence of different levels of residual liquid in successive test pieces, but the final pressure is always related by a fixed pressure difference between the initial and subsequent pressure readings if the leakage rate is zero or constant. This relationship may be experimentally verified by tests on a receptacle containing no leaks, but variable levels of residual liquid.

Beginning with no leak at a given pressure difference between the initial and subsequent pressure readings, increasing leak rates cause a proportionally decreasing difference between the initial and subsequently sensed pressure values.

Variations in the differential pressure readings correspond to the rate of any small leaks, and this fact allows a correction function to be established to correct for the leak. That is, as the leak rate increases, there is a relative reduction in the substantially stabilized fluid pressure relative to the initial sensed pressure, and this effect allows the error in the calculated volume to be computed and applied to offset that error.

It has been discovered that the correction function to be applied to the calculated volume comprises a linear function of the sensed pressure difference of the form:

k₂ (ΔPf)-C_st

The factor k₂ and the constant C_st are determined by experiment with a "dry" receptacle to establish the correcting function.

Detection of a gross leak triggers an alarm, rather than calculation of a corrected volume.

DESCRIPTION OF THE DRAWINGS

FIG. 1 is a diagrammatic representation of net volume measuring apparatus which may be used to practice the method according to the present invention, connected to a sealed automatic transmission containing a small amount of residual transmission fluid.

FIG. 2 is an x-y plot of sensed pressures for three different levels of residual liquid and a constant leakage rate.

FIG. 3 is an x-y plot of pressures for a constant residual volume with different rates of leakage representing three representative test places.

FIG. 4 is a plot of the relationship between the difference between the initial and stabilized sensed pressures and the calculated fill volume.

FIG. 5 is a plot of experimental data on the calculated fill volume and the sensed pressure difference made in order to calculate the correction constant C_st and factor k₂.

DETAILED DESCRIPTION

In the following detailed description, certain specific terminology will be employed for the sake of clarity and a specific embodiment described in accordance with the requirements of 35 USC 112, but it is to be understood that the same is not intended to be limiting and indeed the invention is capable of taking many forms within the scope of the appended claims.

Referring to FIG. 1, the present invention concerns successively measuring the net internal volume of each of a series of test piece receptacles 10, here schematically shown as an automatic transmission, containing an unknown volume of residual liquid V_hr. This is done in this particular application in order to accurately calculate the volume of liquid needed to be added, V_calc to bring the total volume of liquid to a predetermined volume V_ho.

According to the present method, a reference vessel, shown as a tank 12, is first pressurized with a gas, such as a source 14 of pressurized air, to an accurately controlled moderate pressure level, on the order of 4 psig, using a precision regulator 16.

The tank pressure level is monitord by a high accuracy (0.1%) pressure transducer 18, which operates a corresponding electrical signal transmitted to a programmable controller 20, incorporating a microprocessor to receive and record data, perform calculations, and generate control signals, in the manner well known to those skilled in the art.

The reference vessel 12 is able to be controllably placed in communication with transmission 10 by line 22, having a solenoid valve 24 mounted therein, line 22 connected via an adapter 26 to the fill tube 28 of the transmission 10.

The transmission 10 has all its outlet orifices temporarily sealed during the measurement process, such as the vent 30 and the coolant line 32.

During calibration tests to determine certain constants as will be described, one of the orifices, such as the coolant line 32 is controllably sealed with a needle valve 34 to allow establishing a small degree of leakage of progressively increasing flow rates.

The programmable controller 20 generates a control signal at the start of each test to close a solenoid valve 36 to isolate the reference vessel 12, solenoid valve 24 also being previously closed to isolate the vessel 12.

FIG. 2 illustrates a typical cycle for a series of test pieces having varying volumes of residual liquid V_hr +1, V_hr, V_hr -1.

After a short period A, i.e., 5 seconds, to allow stabilization, a series of pressure measurements from the transducer 18 are recorded over a brief period B, i.e., 2 seconds, by the programmable controller 20 to obtain an average preliminary pressure in the reference vessel 12.

The solenoid valve 24 is then opened by a control signal from the programmable controller 20 to place the pressurized reference vessel 12 in communication with the interior space.

The pressurized gas expands immediately into the unpressurized interior spaces of the transmission 10. After a short interval, C, i.e., 3 seconds, a series of pressure readings are recorded over a next brief interval, D, i.e., 2 seconds, to obtain an initial average pressure reading in the system comprised of the interior space of the reference vesssel 12, connecting piping and valves and the transmission 10.

After a much longer interval, E, i.e., 30 seconds, a subsequent series of pressure readings are recorded over period F, to obtain an average pressure value closely approximating the final, stabilized pressure in the system.

According to Boyle's law, the relationship between the preliminary pressure in the reference vessel and connecting piping upstream of the valve 24 and the final stabilized pressure in the entire system, is proportional to the relative volumes, i.e.,

P₁ V₁ =P₂ V₂

If the various constants are known, i.e., the volumes of the reference vessel 12, the piping, etc. the net internal volume of the receptacle 10 can be calculated and thus also the volume V_calc needed to be added to bring the volume of liquid to the desired level V_ho.

The final pressure Pf₂ increases with increasing volumes of residual liquid V_r as shown in FIG. 2.

The reference vessel 12 is thereafter disconnected and repressurized for the next cycle.

Preferably, two reference vessels 12 are alternately used to eliminate the need for a waiting period before initiating the next test.

The programmable controller 20 can then generate suitable control signals to a dispenser unit 38 to dispense the appropriate volume V_calc into the transmission 10 via piping 40, flow meter 42 and solenoid valve 44 and adapter 26. The flow meter 42 generates a digital count which is sent to the programmable controller 20 which shuts off the solenoid valve 44. An air injection separation via line 46 could be employed to break the oil column at the adaptor 26, and the dispenser 38 may suction back the excess oil for reprocessing in order to provide a dripless disconnect.

The above Boyle's law calculation assumes leak free seals and connections. If leaks are present, the final pressure is reduced, as shown in FIG. 3, which depicts pressures for three test pieces 1, 2 and 3, having successively increasing leak rates causing lowered final pressures Pf₂, measured in the interval F.

Thus, if a leak is present of unknown magnitude, the calculation will be erroneous, in that the sensed pressure Pf₂ is falsely lowered correspondingly increasing the magnitude of V_calc.

Thus, an overall formula may be used as follows:

V_calc =k₁ [(Pe/Pf₂ ×V_e)-V_e -V_t +V_ho -V_bs ]-C_st +(k₂ XΔPf)

Where:

V_calc =volume to be added into transmission

k₁ =correction factor (short stabiliztion time)

P_e =average preliminary pressure in reference tank

Pf₂ =average pressure in total system after 30 seconds

V_e =reference tank volume

V_t =volume of pipes of manifolds and adaptor

V_ho =theoretical net volume of the transmission with correct level

V_bs =dry internal volume of transmission (no oil residual)

C_st =correction constant for small leaks

k₂ =correction factor for small leaks

ΔPf=Pf₂ -Pf₁

Pf₁ =average pressure in the system after 3 seconds

k₁ approaches 1.0 and is required where the period E is too short to achieve substantially stabilized pressure in the system.

According to the concept of the present invention, a correction function is developed based on the difference between the initial measured pressure Pf₁, and the subsequent pressure Pf₂ or ΔPf.

FIG. 4 illustrates a plot of empirical data on a receptacle showing that as the differential pressure ΔPf increases, the value of V_calc decreases until at a given pressure differential Pf_o corresponding to a zero leak rate, an accurate or true V_calc is calculated. It is noted that the straight line connecting the points in FIG. 4 comprises a correction function to find V_calc from a given V_calc in the presence of a leak rate and ΔPf_x.

The form of the correction function is:

-C_st +k₂ ΔPf

That is, the algebraic sum of a correction constant, -C_st and the product of a correction factor k₂ and the pressure difference ΔPf is added to a calculated V_calc to yield the true V_calc o.

The values of C_st and k₂ for a given receptacle configuration are arrived at experimentally. That is, a dry transmission is connected in FIG. 1, with a needle valve 34 connected to one of the orifices. After verifying that no leaks exist, a test is run with the needle valve 34 completely closed to determine ΔPf.

One or more additional tests are run with progressive opening of the needle valve 34 to establish controlled leaks of varying rates, with ΔPf readings and calculations ofV_calc by simple Boyle's Law computations.

FIG. 5 illustrates a plot of such values.

At point A, ΔPf is 27 and V_calc is 7.303.

At point B, ΔPf is 14 and V_calc is 9.303, so a correction of ΔV_calc equal to 2 is required.

Equation at point A is:

correction value=ΔV_calc =0=k₂ 27 -C_st

At point B:

ΔV_calc =+2=k₂ 14-C_st

Solving these equations simultaneously gives:

k₂ =0.154

C_st =4.154

The test is again run to verify that V_calc will remain the same for any setting of the needle valve 34.

If there is a variation, the slope can be corrected depending on the test results. For example, if the correction is too high with increasing leak rates, the slope of the curve must be lowered slightly, as illustrated in FIG. 5.

That is, a new point C is plotted:

ΔV_calc =2; ΔPf=13

Resolving the equations for A and C:

k₂ =0.143 C_st =3.860

To fully verify, the test can be run after adding exactly one liter to ensure repeatability of 0.3 liters. Verifying the fill level in the transmission after adding V_calc should be done.

If the oil levels resulting from adding the corrected calculated fill volumes of oil V_calc are mainly too high, decreasing the assumed theoretical V_ho value and vice versa will if too low make closer results achievable by thereby determining empirically a more precise value for subsequent use in testing of receptacles in the series being tested.

Referring again to FIG. 1, if a gross leak is detected by detecting a decline of the pressure difference ΔPf below a predetermined level, a fault alarm 50 may be triggered and the calculation disabled. This is because the later measured pressure Pf₂ reduced by the presence of a leak, and the greater the leak the greater the reduction. However since Pf₂ is normally greater than Pf₁ (see FIG. 2), a reduction in Pf₂ results in a reduction in ΔPf. Thus, an operator may inspect the seals, etc., or merely add a predetermined volume of fluid manually.

Claims

1. A method of determining the net internal volume of each of a series of test piece receptacles each containing a variable but unknown volume of liquid, the volume of which is to be subtracted from the dry internal volume of said receptacles to arrive at said net internal volume, to thereby enable calculation of the volume of liquid to be added to bring the level of liquid in each piece test piece receptacle to a predetermined level, comprising the steps of:

pressurizing a reference vessel having an interior space of known volume with a gas and thereafter preliminarily sensing the pressure level in said pressurized reference vessel;

placing said pressurized vessel in communication with the interior space of a selected one of said receptacles by means of fluid passages of known volume so as to allow expansion of said gas in said reference vessel thereinto;

initially sensing the gas pressure in a system comprising said connected reference vessel, said fluid passages and said selected receptacle shortly after placing the reference vessel in communication with the selected receptacle but substantially before said pressure stabilizes therein;

subsequently sensing the gas pressure in said system again after said gas pressure has substantially stabilized therein;

determining the difference between said initially and subsequently sensed gas pressures;

calculating the net internal volume of said selected receptacle from the proportion of the preliminarily sensed pressure level in said reference vessel to the subsequently sensed pressure level in said system and the value of said dry internal volume of said receptacle;

correcting the calculated value of said net internal volume of the selected receptacle by reducing said calculated value by a correction function inversely related to the difference between said initially and subsequently sensed pressure levels;

repeating each of said steps for each successive test piece receptacle.

2. The method according to claim 1 wherein said correction function is of the form:

k₂ ΔPf -C_st

where: k₂ is a correction factor; ΔPf is the difference between said initially and subsequently sensed pressures; and, C_st is a constant equal to k₂ ΔPf for the ΔPf produced with no leak present.

3. The method according to claim 2 where k₂ and C_st are determined by testing a dry reference receptacle to determine the correction function by introducing one or more leak rates from said reference receptacle.

4. The method according to claim 2 further comprising the step of calculating the volume of liquid to be added to said selected receptacle by subtracting the net internal volume of said selected receptacle containing said predetermined level of liquid from said calculated volume.

5. The method according to claim 4 wherein each of said receptacles is substantially identical, said method further comprising assuming a theorical net internal volume of said selected receptacle when it has been filled to the predetermined level and subsequently adjusting the value of said assumed net internal volume after measuring the actual liquid level in said selected receptacle resulting after adding said calculated volume of liquid, to thereby determine empirically the precise net internal volume value for subsequent testing of receptacles in said series.

6. The method according to claim 1 further including generating an alarm signal whenever the leak rate exceeds a predetermined rate as determined by a detection of a decline of said pressure difference below a predetermined level.

7. A method of filling an automatic transmission with a desired volume of fluid, said transmission having interior spaces of a known internal dry volume, with an unknown residual volume of fluid contained therein to achieve a predetermined fluid level in said transmission, comprising the steps of:

pressurizing with a gas a reference vessel to an accurately controlled pressure level;

subsequently placing said pressurized vessel in communication with the interior spaces of said transmission after sealing the outlets from said transmission so as to allow expansion of said gas through a system including said reference vessel and the interior spaces of said sealed transmission;

initially reading the pressure in said system shortly after expansion is initiated and subsequently again after said pressure substantially stabilizes;

determining the net volume of said transmission interior spaces, i.e., the internal dry volume of said interior spaces less the volume of said residual fluid from the change in pressure from the predetermined level in said reference vessel to the stabilized pressure level;

calculating the fill volume of fluid to be added by subtracting the net volume from the known internal dry volume to find the residual volume, and subtracting the residual volume from the predetermined desired volume;

correcting the calculated fill volume by a correction function of the form:

-C_st +k₂ ΔPf

where C_st is a constant correction, k₂ is a correction factor, and ΔPf is the difference between said initially and subsequent read pressures;

deriving the values of C_st and k₂ empirically by testing a sample transmission configured identically to said transmission to be filled to simulate various small leaks and determining the effect such small leakage has on ΔP and the calculated volume of the transmission internal spaces.